t/81_hosts.t fails when OpenDNS is configured as DNS provider

[Christian Walde]

On Tue, 01 Oct 2013 21:26:01 +0200, Marc Lehmann <schmorp at schmorp.de>  
wrote:

> On Tue, Oct 01, 2013 at 02:21:22PM +0200, Christian Walde  
> <walde.christian at gmail.com> wrote:
>> The test in t/81_hosts.t attempts to verify that PERL_ANYEVENT_HOSTS
>> is used correctly, by trying to resolve an invalid host, and
>> expecting it to fall back. When OpenDNS is configured as a DNS
>> provider however, it responds with the IP 67.215.65.132 to requests
>> of non-existant or invalid hosts.
>
> AnyEvent requires a working DNS server - if the test fails with opendns,
> then your dns configuration is indeed not working properly, and failure  
> is expected.
>
>> I'm willing to provide a patch for this. Which resolution for this
>> would you prefer?
>
> This is a problem outside the scope of AnyEvent, really. If DNS is that
> badly broken, then a test failure is the least of your problems.

Test fails are meant to indicate a failure with the distribution itself,  
not
incompatibilities with a broken environment. As such the current test FAIL
is a mis-signal, since there is no actual fault in the AE code.

If the incompatibility is of a bad enough scale, then this should be
indicated with an NA result; if it is small enough to merely impede part of
overall functionality, then that test should be skipped.

Since this kind of breakage is easily detected, I am willing to provide you
a patch to implement either kind of signaling mechanism, as by your choice.

> Note that you can still install AnyEvent just fine - the test merely
> signals that you will not correct dns resolution.

I'm aware of the option, and while i am a user, this issue is not impeding
my use since the issue occurs on a smoke tester. Still, thanks for the good
intentions in pointing this out.

-- 
With regards,
Christian Walde